No abstract

Summary N 6 -methyladenosine (m 6 A), the most prevalent internal RNA modification on mammalian messenger RNAs (mRNAs), regulates fates and functions of modified transcripts through m 6 A-specific binding proteins 1 – 5 . m 6 A is abundant in the nervous system and modulates various neural functions 6 – 11 . While m 6 A marks groups of mRNAs for coordinated degradation in various physiological processes 12 – 15 , the relevance of m 6 A in mRNA translation remains largely unknown in vivo . Here we show that, through its binding protein Ythdf1, m 6 A promotes protein synthesis of target transcripts in response to neuronal stimuli in the adult mouse hippocampus, thereby facilitating learning and memory. Mice with genetic deletion of Ythdf1 ( Ythdf1 -KO) exhibit learning and memory defects as well as impaired hippocampal synaptic transmission and long-term potentiation. Ythdf1 re-expression in the hippocampus of adult Ythdf1 -KO mice rescues behavioral and synaptic defects, while hippocampus-specific acute knockdown of Ythdf1 or Mettl3 , the catalytic component of m 6 A methyltransferase complex, recapitulates the hippocampal deficiency. Transcriptome-wide mapping of Ythdf1 binding sites and m 6 A sites on hippocampal mRNAs uncovered key neuronal genes. Nascent protein labeling and tether reporter assays in hippocampal neurons revealed that Ythdf1 enhances protein synthesis in a neuronal-stimulus-dependent manner. Collectively, our results uncover a pathway of mRNA m 6 A methylation in learning and memory, which is mediated through Ythdf1 in response to stimuli.

Abstract-In this work, we consider the problem of pedestrian detection in natural scenes. Intuitively, instances of pedestrians with different spatial scales may exhibit dramatically different features. Thus, large variance in instance scales, which results in undesirable large intra-category variance in features, may severely hurt the performance of modern object instance detection methods. We argue that this issue can be substantially alleviated by the divide-and-conquer philosophy. Taking pedestrian detection as an example, we illustrate how we can leverage this philosophy to develop a Scale-Aware Fast R-CNN (SAF R-CNN) framework. The model introduces multiple built-in subnetworks which detect pedestrians with scales from disjoint ranges. Outputs from all the sub-networks are then adaptively combined to generate the final detection results that are shown to be robust to large variance in instance scales, via a gate function defined over the sizes of object proposals. Extensive evaluations on several challenging pedestrian detection datasets well demonstrate the effectiveness of the proposed SAF R-CNN. Particularly, our method achieves state-of-the-art performance on Caltech [8], INRIA [5], and ETH [9], and obtains competitive results on KITTI [11].

Detecting small objects is notoriously challenging due to their low resolution and noisy representation. Existing object detection pipelines usually detect small objects through learning representations of all the objects at multiple scales. However, the performance gain of such ad hoc architectures is usually limited to pay off the computational cost. In this work, we address the small object detection problem by developing a single architecture that internally lifts representations of small objects to "super-resolved" ones, achieving similar characteristics as large objects and thus more discriminative for detection. For this purpose, we propose a new Perceptual Generative Adversarial Network (Perceptual GAN) model that improves small object detection through narrowing representation difference of small objects from the large ones. Specifically, its generator learns to transfer perceived poor representations of the small objects to super-resolved ones that are similar enough to real large objects to fool a competing discriminator. Meanwhile its discriminator competes with the generator to identify the generated representation and imposes an additional perceptual requirement -generated representations of small objects must be beneficial for detection purpose -on the generator. Extensive evaluations on the challenging Tsinghua-Tencent 100K [45] and the Caltech [9] benchmark well demonstrate the superiority of Perceptual GAN in detecting small objects, including traffic signs and pedestrians, over well-established state-of-the-arts.

ObjectivesTo investigate superiority of a telerehabilitation programme for COVID-19 (TERECO) over no rehabilitation with regard to exercise capacity, lower limb muscle strength (LMS), pulmonary function, health-related quality of life (HRQOL) and dyspnoea.DesignParallel-group randomised controlled trial with 1:1 block randomisation.SettingThree major hospitals from Jiangsu and Hubei provinces, China.Participants120 formerly hospitalised COVID-19 survivors with remaining dyspnoea complaints were randomised with 61 allocated to control and 59 to TERECO.InterventionUnsupervised home-based 6-week exercise programme comprising breathing control and thoracic expansion, aerobic exercise and LMS exercise, delivered via smartphone, and remotely monitored with heart rate telemetry.OutcomesPrimary outcome was 6 min walking distance (6MWD) in metres. Secondary outcomes were squat time in seconds; pulmonary function assessed by spirometry; HRQOL measured with Short Form Health Survey-12 (SF-12) and mMRC-dyspnoea. Outcomes were assessed at 6 weeks (post-treatment) and 28 weeks (follow-up).ResultsAdjusted between-group difference in change in 6MWD was 65.45 m (95% CI 43.8 to 87.1; p<0.001) at post-treatment and 68.62 m (95% CI 46.39 to 90.85; p<0.001) at follow-up. Treatment effects for LMS were 20.12 s (95% CI 12.34 to 27.9; p<0.001) post-treatment and 22.23 s (95% CI 14.24 to 30.21; p<0.001) at follow-up. No group differences were found for lung function except post-treatment maximum voluntary ventilation. Increase in SF-12 physical component was greater in the TERECO group with treatment effects estimated as 3.79 (95% CI 1.24 to 6.35; p=0.004) at post-treatment and 2.69 (95% CI 0.06 to 5.32; p=0.045) at follow-up.ConclusionsThis trial demonstrated superiority of TERECO over no rehabilitation for 6MWD, LMS, and physical HRQOL.Trial registration numberChiCTR2000031834.